Anaerobic biological treatment has been in existence since the 1910's when it was first used for the treatment of municipal sewage sludge. Over the last twenty-five years, a number of new anaerobic processes have been developed for the treatment of industrial wastes, as well as for conventional sludge digestion.
Industrial waste pretreatment is becoming commonly viewed as a cost-savings measure by generating industries, in that it often costs less to pretreat wastes than to pay municipal sewer usage fees. Anaerobic processes are now being considered by many industries as an attractive alternative for industrial waste pretreatment.
The pioneering work on the anaerobic filter was performed at Stanford University by Young for his doctoral research under McCarty, and was presented in detail in 1969. [Young, J. C., and P. L. McCarty "The anaerobic filter for waste treatment." JWPCF, 41, 5, R160-R173 (1969)].
The first anaerobic filters were described by Young as completely submerged, 12-liter, laboratory-scale reactors which were filled with 1.0 to 1.5 inch quartzite stone (bed porosity of 0.42). The reactors were continuously-fed in a upflow mode at 25.degree. C.
Two substrates were used, a mixture of proteins and carbohydrates, and a mixture of acetic and propionic acids. Waste strengths ranged from 1,500 mg/L to 6,000 mg/L at hydraulic retention times (HRTs) ranging from 4.5 hrs to 72 hrs, resulting in organic loadings ranging from 0.43 g COD/L/day to 3.40 g COD/L/day. Treatment success was based primarily on COD removals, based on measured influent and effluent CODs. COD removals ranged from a high of 93.4% at the 72 hr HRT at a loading of 0.43 g COD/L/day, to a low of 36.7% removal at the 4.5 hr HRT at a loading of 3.4 g COD/L/day.
From the initial research on the anaerobic filter, Young and McCarty stated several observations and advantages of the anaerobic filter as compared to other anaerobic treatment systems including:
1. The anaerobic filter is ideal for the treatment of soluble waste streams. PA1 2. Biological solids accumulate in the anaerobic filter leading to long solids retention times (SRTs), and low effluent total suspended solids (TSS). PA1 3. Because of the long SRTs possible, dilute wastes can be treated successfully at nominal temperatures (&lt;37.degree. C.).
Shortly after Young and McCarty's early work on the anaerobic filter, Plummer applied the anaerobic filter treatment process to an actual food processing waste, consisting mainly of carbohydrates. [Plummer, A. H., Jr., Maline, J. F., Jr., and W. W. Ekenfelder, Jr. "Stabilization of a low-solids carbohydrate waste by an anaerobic submerged filter." Proceedings of the 23rd Industrial Waste Conference, Purdue University, Lafayette, Ind. (1968)]. Instead of using rock packing media, Plummer used a plastic ring and saddle media. This provided a much higher filter bed porosity (0.70 porosity vs. 0.42) than was the case for Young's rock-filled filter. The higher bed porosity left more physical space in the reactor for the retention of anaerobic biomass.
Plummer's filters were operated mesophilically at organic loadings of 1.6 g COD/L/day to 10.3 g COD/L/day at HRTs ranging from 13 hrs to 83 hrs, resulting in COD removals ranging from 41 to 93.5%. To avoid the possibility of liquid short-circuiting, Plummer suggested effluent recycling.
The anaerobic filter process was applied to a pharmaceutical waste by Jennett and Dennis in 1975. [Jennett, J. C., and N. D. Dennis, Jr. "Anaerobic filter treatment of pharmaceutical waste." JWPCF, 47, 1, 104-121 (1975)]. The pharmaceutical waste was low in suspended solids and had an average COD of 16,000 mg/l. The reactors were filled with 1.0 to 1.5 inch gravel, and had a 14 liter empty bed volume. HRTs of 12 hrs to 48 hrs were studied at 37.degree. C. and applied organic loadings ranging from 0.2 g COD/L/day to 3.5 g COD/L/day. COD removal efficiencies ranged from 94 to 98%. An important observation made by Jennet and Dennis was that bacteria collected on and between the interstitial spaces of the gravel media.
Chain and DeWalle utilized an anaerobic filter for the treatment of acidic landfill leachate which had a pH of 5.4 and a COD of 54,000 mg/L. [Chian, E. S. K., and DeWalle, F. B., "Treatment of High Strength Acidic Wastewaters with a Completely Mixed Anaerobic Filter." Water Research, Vol. 11 (1977)]. They practiced effluent liquid recycle in order to help neutralize the pH of the acidic leachate. They also used plastic media which provided for a high filter bed porosity of 0.94.
The anaerobic filter was used to treat shellfish processing wastewaters by Hudson. [Hudson, J. W., Pohland, F. G., and Pendergrass, R. P. "Anaerobic packed column treatment of shellfish processing wastewaters." Proceedings of the 34th Industrial Waste Conference, Purdue University, Lafayette, Ind. (1984)]. Two different types of packing media were used, including readily-available oyster shells and stone media, resulting in bed porosities of 0.82 and 0.53, respectively. This was an important study which illustrated how filter bed porosity may affect wastewater treatment. The oyster shell media filter with the higher bed porosity provided superior treatment as compared to the stone media, with COD removals of 81%.
In 1980, Dague reported on the use of anaerobic filters to treat a high-strength grain processing waste. [Dague, R. R. "Principles of anaerobic filter design." Presented at the 26th annual Great Plains Wastewater Design Conference, Omaha, Nebr., Mar. 23, 1982.] Two temperatures were utilized, 22.degree. C. and 35.degree. C., at applied organic loadings of 2.4 g COD/L/day. COD removal efficiencies were 75% at 22.degree. C., and 90% at 35.degree. C. The organic loading was increased to 5.6 g COD/L/day for the 35.degree. C. filter, but treatment was unstable due to pH fluctuations in the raw waste.
Stevens and van den Berg used the anaerobic filter process to treat food processing wastes at 37.degree. C. [Stevens, T. G., and L. van den Berg "Anaerobic treatment of food processing wastes using a fixed-film reactor." Proceedings of the 36th Industrial Waste Conference, Purdue University, Lafayette, Ind. (1981)]. Two wastes were utilized including a bean-blanching wastewater, and a caustic tomato-peeling wastewater, which was neutralized to a pH of 9.5 prior to anaerobic treatment.
The single filter was operated in the upflow mode with the bean-blanching waste and in the downflow mode with the bean-blanching waste and the tomato-peeling waste. Little change in treatment was observed when the filter was instantaneously switched from the bean-blanching waste to the tomato-peeling waste.
In 1984, Guiot and van den Berg described a modified anaerobic filter, termed the upflow blanket filter (UBF), or hybrid filter. [Guiot, S. R. and L. van den Berg "Dynamic performance of an anaerobic reactor combining an upflow sludge blanket and a filter for the treatment of sugar waste." Proceedings of the 39th Industrial Waste Conference, Purdue University, Lafayette, Ind. 1984]. In their design, the bottom two-thirds of the reactor consisted of an open space where a sludge blanket formed. The top one-third of the reactor contained conventional plastic media. They tested a synthetic waste consisting mainly of sucrose in which applied organic loadings of up to 22 g COD/L//day resulted in 95% soluble COD removal efficiencies.
Many more studies have been conducted using anaerobic filters. The anaerobic filter has been applied mainly to soluble waste streams using a variety of synthetic as well as actual industrial wastes.
There are no known two-stage anaerobic treatment processes in which a thermophilic anaerobic filter is connected in series to a mesophilic anaerobic filter. The majority of the relevant literature deals with two-stage systems designed for enhanced phase optimization.
In 1971, Pohland and Ghosh first proposed a two-phase system for the separation of the acidogenic and methanogenic phases of anaerobic treatment. [Pohland, F. G., and S. Ghosh "Developments in anaerobic stabilization of organic wastes--the two-phase concept." Env. Letters, 1, 4, 255-266 (1971)].
Their system consisted of two completely-mixed reactors connected in series for waste stabilization. They noted that by separating the acid-forming organisms from the methanogenic organisms, optimal growth environments could be maintained for each population. By the use of kinetic control using the appropriate dilution rates, they proposed that the rapid-growing acidogens would be predominate in the first stage, forming volatile acids. The slower-growing methanogens would be washed out of the first stage and predominate in the second stage where they could convert volatile acids produced in the first stage to methane gas. They noted that the key to successful treatment was near-complete phase separation.
In 1985, Ghosh compared the two-stage process to single-stage digestion for the anaerobic treatment of sewage sludge. [Ghosh, S. "Improved sludge gasification by two-phase anaerobic digestion." J.Env. Eng,. 113, 1265-1284 (1985)]. Continuously-mixed reactors were used, and temperature variations of mesophilic to mesophilic, mesophilic to thermophilic, and thermophilic to thermophilic were applied for the first and second stages, respectively. Ghosh noted that the two-stage systems showed superior performance when compared to the single-stage system, especially at high organic loadings and low HRTs.
Tanaka and Matsuo treated a dilute milk waste stream using a two-stage system which consisted of a continuously-mixed reactor connected in series to a methanogenic anaerobic filter at 37.degree. C. [Tanaka, S. and T. Matsuo "Treatment characteristics of the two phase anaerobic digestion system using an upflow filter." Wat.Sci.Tech., 18, 217-224 (1986)]. At a system HRT of 4.4 days, 92% COD removals were achieved at a low organic loading of 1.5 g COD/L/day. They observed more successful phase separation when the acidogenic continuously-mixed reactor was operated at a 1-day HRT rather than a 2-day HRT. In an analysis of the acidogenic effluent from the first stage, it was observed that carbohydrates were more readily degraded than the proteins or lipids present in the milk waste stream.
In 1986, Lo and Liao were the first to utilize the two-stage process for an agricultural waste, screened dairy manure. [Lo, K. V., and P. H. Liao "Thermophilic anaerobic digestion of screened dairy manure using a two-phase process." Energy in Agriculture, 5, 249-255 (1986)]. They compared one-stage and two-stage systems at 55.degree. C. The two-stage system consisted of a 6-liter completely-mixed reactor operated at HRTs of 1.3 to 1.5 days connected to a 4-liter anaerobic filter operated at a 1-day HRT. The single-stage system consisted of a 4-liter anaerobic filter operated at a 1-day HRT. Better performance was observed in the single-stage system for the screened dairy manure.
In 1990, Hanaki and others compared single-stage and two-stage anaerobic treatment of an oily cafeteria wastewater at 20.degree. C. [Hanaki, K. Matsuo, T., and K. Kumazaki "Treatment of oily cafeteria wastewater by single-phase and two phase anaerobic filter." Wat. Sci. Tech., 22, 3/4, 299-306 (1990)]. Similar to previous research, the two stage system consisted of a completely-mixed reactor connected in series to an anaerobic filter. The cafeteria wastewater contained approximately 30% lipids, and had a COD of 1,300 mg/L to 2,500 mg/L. Slightly better COD removals were observed in the single-stage filter as compared to the two-stage system.
Aoki and Kawase reported on the use of a two-stage process at a thermophilic temperature to digest sewage sludge. [Aoki, N., and M. Kawase "Development of high-performance thermophilic two-phase digestion process." Wat.Sci.Tech., 23, 1147-1156 (1991)]. Their research was slightly different in that a thermal conditioning pretreatment step was applied at 90.degree. C. for one hour using a proteolytic enzyme. The two-stage system consisted of a 70.degree. C. completely-mixed acidogenic reactor connected to a 55.degree. C. anaerobic filter. Their system obtained a 58% volatile solids reduction at a system HRT of 3.7 days.
Other reports in the literature deal with multi-stage designs which were not developed for phase optimization.
In 1973, El-Shafie and Bloodgood described an experiment in which six anaerobic filters were connected in series for the treatment of Metrecal (vanilla flavor) at 30.degree. C. [El-Shafie, A., and D. E. Bloodgood "Anaerobic treatment in a multiple upflow filter system." JWPCF, 45, 11, 2345-2357 (1973)].
In their system, the six reactors were filled with 1 to 1.5 inch gravel media and had a working volume of 2.6 liters each. The Metrecal waste stream had a COD of 10,000 mg/l, and was applied at the rate of 1 liter per hour, resulting in an effective load on the first anaerobic filter of 41 g COD/L/day. The retention time in each of the six filters was three hours, resulting in a system HRT of 18 hours for the combined six filters. System COD removals averaged 76. El-Shafie and Bloodgood observed that there was an exponential decrease in biological activity from the first to the last filter.
Howerton and Young investigated a unique two-stage cyclic operation of anaerobic filters using a synthetic alcohol stillage waste comprised mainly of ethanol and sucrose. [Howerton, D. F., and J. D. Young "Two-stage cyclic operation of anaerobic filters." JWPCF, 59, 8, 788,794 (1987)].
In their system, two 370-liter anaerobic filters were connected in series, with the first reactor termed the lead reactor, and the second reactor termed the follow reactor. As a part of their study, after 136 days of continuous operation of the filters at 30.degree. C., the waste flow was reversed, with the follow reactor becoming the lead reactor. At organic loadings of 4 g COD/L/day and 8 g COD/L/day, using system HRTs of 36 hrs and 18 hrs, COD removals of 98% to 99% were observed.
At Iowa State University, Harris conducted a comparative study of mesophilic and thermophilic anaerobic filters for his doctoral research under Dague. [Harris, W. L. Comparative performance of anaerobic filters at mesophilic and thermophilic temperatures." Doctoral dissertation, Library, Iowa State University (1992)].
The laboratory-scale anaerobic filters had clean-bed volumes of 16.8 liters. Non-fat dried milk was used as the substrate. The mesophilic and thermophilic filters were operated at 35.degree. C. and 56.degree. C., respectively. Harris observed that the thermophilic reactors produced a lower quality effluent than the mesophilic reactors at high organic loadings.